Update today at 4pm EST time.

iscDNase-seq

Workflow for iscDNase-seq analysis

GEO: GSE136987 (private)

Figure 3 plot generation workflow

Assumptions:

• Files are locally accessible
• Access to Biowolf high performance computing resources Both of these assumptions will have to be corrected for at some point with analysis conducted starting with GEO over less sophisticated servers.

Building tools

HOME=`pwd`

cd ~/src/Figure3
g++ -o generateRPBMBasedSummary OftenUsedOperatLib.cpp bed.cpp ucsc.cpp TypeDefBase.h Matrix.cpp SequenceTransform.cpp expression.cpp UCSCBEDOperator.cpp generateRPBMBasedSummary_v2.cpp 

if [[ ! -f bigWigToWig ]]; then
    wget http://hgdownload.cse.ucsc.edu/admin/exe/linux.x86_64/bigWigToWig
fi

if [[ ! -f hg19ToHg18.over.chain.gz ]]; then
    wget http://hgdownload.soe.ucsc.edu/goldenPath/hg19/liftOver/hg19ToHg18.over.chain.gz
fi

if [[ ! -f rpkmforgenes.py ]]; then
	wget http://sandberg.cmb.ki.se/media/data/rnaseq/rpkmforgenes.py
fi

chmod 777 generateRPBMBasedSummary bigWigToWig hg19ToHg18.over.chain.gz rpkmforgenes.py

Acquiring data

cd ~/data/Figure3
mkdir -p sc_atac i_dnase b_atac b_dnase cluster
mkdir -p sc_atac/bedfiles i_dnase/bedfiles b_atac/bedfiles b_dnase/bedfiles

cd sc_atac
bash ~/src/Figure3/get_sc_atac.sh 
 
cd ../i_dnase
bash ~/src/Figure3/get_i_dnase.sh 

cd ../b_atac
bash ~/src/Figure3/get_b_atac.sh 

cd ../b_dnase
bash ~/src/Figure3/get_b_dnase.sh

Finding peaks

cd ..
mkdir -p macs peaks
cd macs 
bash ~/src/Figure3/make_macs_swarm.sh 0 macs.swarm
swarm -f macs.swarm -g 32 --module macs --gres=lscratch:64

bash ~/src/Figure3/make_macs_swarm.sh 1 macs.sub.swarm
swarm -f macs.sub.swarm -g 32 --module macs --gres=lscratch:64

cd ../peaks 
bash ~/src/Figure3/get_peaks.sh 
bash ~/src/Figure3/get_blacklist.sh 
bash ~/src/Figure3/merge_peaks.sh

3d

bash ~/src/Figure3/make_ucsc_swarm.sh /data/perriejv/genome_len/hg18_chrlen.txt ucsc.swarm
swarm -f ucsc.swarm -g 32 --gres=lscratch:64

mkdir -p inter_peaks rand_peaks anno_peaks

cd ..

Viz

source /data/$USER/conda/etc/profile.d/conda.sh
conda activate base
conda activate general

3a/b

Rscript ~/src/Figure3/assay_similarties.R

3c

Rscript ~/src/Figure3/find_uniq_inter_peaks.R

3e

# navigate to http://hgdownload.cse.ucsc.edu/goldenpath/hg18/phastCons17way/ and make a links to each chromosome 
mkdir -p phastcons rna_seq 
wget -i filepaths
gunzip *.gz
module load bedtools
python ~/src/Figure3/make_rand_peaks.py
python ~/src/Figure3/find_conserved_profile.py

3f

# add bigWig signal of all reads for the following cell types: 
# Monocyte: https://www.encodeproject.org/experiments/ENCSR905LVO/, ENCFF706XVQ
# B cell: https://www.encodeproject.org/experiments/ENCSR449GLL/, ENCFF517AYD
# T cell: https://www.encodeproject.org/experiments/ENCSR545MEZ/, ENCFF947QSD
# Mar. 2006 (NCBI36/hg18), Other RefSeq, xenoRefGene
cd ../rna_seq
wget -i filepaths
module load python
python ~/src/Figure3/rpkmforgenes.py -i ENCFF517AYD.bigWig.hg18.bed  ENCFF706XVQ.bigWig.hg18.bed  ENCFF947QSD.bigWig.hg18.bed -readcount -a hgTables.txt -o rna_seq_gene_expression.txt -fulltranscript -rmnameoverlap -p 20

cd ../peaks
bash ~/src/Figure3/make_anno.sh ~/data/Figure3/rna_seq/rna_seq_gene_expression.txt
cd anno_peaks
swarm -f anno.swarm -g 16 -t 2 --module homer
python ~/src/Figure3/gene_expr.py

Software tools

1. chromVAR (Schep et al., 2017)
2. HOMER (Heinz et al., 2010): https://hpc.nih.gov/apps/homer.html
3. CrossMap (Zhao et al., 2014): https://hpc.nih.gov/apps/crossmap.html
4. MACS2 (Zhang et al., 2009): https://hpc.nih.gov/apps/macs.html